As is generally known, virtualization and NAS (network-attached storage) are continuing to emerge as viable storage arrangements. Tiered storage has generally been giving way to a more complex system of storage stacks, wherein each stack includes isolated “storage layers”.
In a conventional arrangement as just described, each storage layer manages its own cache independently. Thus, for a single I/O (input/output) path, there can easily result a condition of data duplication across multiple storage layers. Efficient cache management across one or more storage stacks thus becomes a highly elusive task.